Another Atari Lynx II repair thread

[bobmoo79]

I have a dead Lynx II and sadly I have been less than methodical in trying to fix this thing owing to the fact I'm so busy on multiple projects - I've just been hopping between them whenever I get a chance. Never a good thing.

 

I hope that somebody here can offer some advice on what I could try going forward.

 

Symptoms:

It does not respond when fully charged Eneloop batteries and a cartridge are installed and the 'on' button is pressed. Completed dead.

(I initially only tested with batteries)

 

Repair log:

1. I changed the MOSFET and recapped it. No change. Still dead when 'on' is pressed.
2. I rechecked all caps for the correct value and orientation in circuit. All are fine.
3. I have bridged the connection between the negative battery terminal and the earphone socket with a game installed and the backlight comes on but the screen is black and there's no sound.
4. I added 5v to C41. Same result as above.
5. I checked D13 in circuit and it was neither short nor open. I removed it and checked it in a simple circuit. It is functioning properly. Reinstalled C13.
6. I checked D9, D10, D11 and they all checked out as expected for diodes, BUT there wasn't power jack inserted and I didn't realise that in this situation D10 should appear as a short!
7. I later sprayed contact cleaner into the power jack and installed/removed the jack repeatedly. The behaviour of D10 with jack inserted Vs not inserted is now correct, so there was definitely a problem here. Now fixed.
8. Tried Lynx on batteries and with a power supply to see what happens. No change. nothing happens when 'on' is pressed.
9. I removed the controller membrane circuit and tested the operation of the 'on' button using a DVM. It works.
10. I noticed that the end of the membrane connector was worn, possibly due to me installing and removing it multiple times, so I cut off the last 2mm or so.
11. I found the schematic for the power section online and checked all resistor values with a DVM. All check out ok, except R47, which measures at 23ohm instead of 4.7k. Problem? I'll check later.
12. I then rechecked the connections on Q12 and realised that the tab (Drain) isn't connected properly. Doh!!! Why didn't I check that after step 1?? Definite problem, easy fix.

 

Next actions:

1. Reflow the MOSFET Drain connection.
2. Remove R47 from circuit and check value again. Replace if not as expected.
3. Remove Q7 from circuit and test with a DVM.
4. Check voltages on U6 pins to see if they are as expected when the Lynx is in its OFF state. I read somewhere the voltages should be as shown below (I'm not sure why considering the supply is 9V):

 

Pin 1 = 12V

Pin 2 = 0V

Pin 3 = 0V

Pin 4 = 12V

Pin 5 = 12V

Pin 6 = 0V

Pin 7 = 0V (Supply)

Pin 8 = 0V

Pin 9 = 12V

Pin 10 = 12V

Pin 11 = 0V

Pin 12 = 0V

Pin 13 = 12V

Pin 14 = 12V (Supply)

 

 

5. Remove Q8 from circuit and test with a DVM.

 

 

I'll test again after each step.

 

 

If anybody has anything to add then please feel free to chip in!

 

 

[bobmoo79]

Updated progress in 1st post. 

[+Stephen Moss]

22 hours ago, bobmoo79 said:

Check voltages on U6 pins to see if they are as expected when the Lynx is in its OFF state. I read somewhere the voltages should be as shown below (I'm not sure why considering the supply is 9V):

It is the way cheap game system power supplies work, then  the Lynx is on more current will be consumed and the voltage will drop due to losses in the power unit transformer.

The listed voltages for Q6 are correct, press the on button, in the on state pins 10, 11, 12 & 13 should reverse state, pin 7 & 14 should remain as they are and the other pins should read some intermediate value as pins 1 & 2 along with the associated R, C and transistor create an oscillator. 

 

Make sure you are using a lynx 2 schematic not the lynx 1, although the design of the power stage is the same that cannot be said for all the component numbering, although most are the same there are some differences so if you have the wrong schematic you may be checking the wrong components. 

[bobmoo79]

Thanks for the help. I was using the Lynx II power schematic posted on AA in 2015 but after your post found the Hayato schematic so will use that one from now on.

 

I checked voltages on U6. They weren't as posted previously - the 12v were actually just under 9v, which I guessed was correct given that the supply is 9v.

 

 

I re-flowed the MOSFET Drain and confirmed connectivity but no improvement.

 

I removed D13 again and confirmed it is definitely working as expected in my test circuit. I carefully replaced it in the circuit and noticed that I have 9v at BOTH sides of the diode which seemed odd so quickly turned off the system.

 

To be honest I'm not sure which reference point I should be using as the 0v on this test. I tihnk I used GND on the dc power jack, but looking at the Hayato schematic, perhaps I should have used TP19, which is marked as 'zero' (presumably 0v)?? My electronics skills are very rusty!!

 

 

I removed R47 from the circuit and confirmed it is indeed a 4.7kohm resistor as it is supposed to be and not damaged. As it measures much less in circuit (around 23ohm) I was suspicious and checked for continuity across the bare pads on the PCB and it shows a short. So the 5v rail and the GND rail appear to be shorted. I decided to investigate why this is the case.

 

I looked at the power supply circuit to see which components could bridge 5v and GND.

I removed 1 leg of both C39 and C41 and confirmed those caps are not shorted.

I checked C36 and R67 - both ok.

 

To be honest I don't fully understand this circuit so I'm scratching around a bit trying to find anything that is obviously not working.

 

Next actions:

1. Check the operation of U6 as in Stephen's post above.
2. Continue checking components in the power section that could bridge 5v to GND (e.g. C35, C44 etc).
3. Remove Q7 from circuit and test with a DVM.
4. Remove Q8 from circuit and test with a DVM.
5. Remove U6 and try in a test circuit.

 

 

Questions for others:

 

• Which gnd/0V point should I use when measuring the voltage at D13?
• Where is TP15 located? I can't find it anywhere? Is it under the metal shield? (I reinstalled that after the cap replacement so won't see it if it's there).
• Has anybody every noticed an apparent short across R47 before?
• Is there any way to isolate the power circuit from the rest of the PCB while I run my tests?
• Somebody mentioned fuses on the PCB that allow different sections of the circuit to be tested but I don't see them on the schematic. Can anybody explain where to find the fuses and which part of circuit is disabled by removing them?

 

 

 

 

[+karri]

2 hours ago, bobmoo79 said:

• Somebody mentioned fuses on the PCB that allow different sections of the circuit to be tested but I don't see them on the schematic. Can anybody explain where to find the fuses and which part of circuit is disabled by removing them?

There is some black zero ohm resistors linking three or four islands together. Right now I do not remember if it was a Lynx I or II model where I tried to find out which chip had acted as a fuse. Removing them only makes sense if the 5V is shorted to ground and you cannot find out the broken part.

[+Stephen Moss]

21 hours ago, bobmoo79 said:

I removed D13 again and confirmed it is definitely working as expected in my test circuit. I carefully replaced it in the circuit and noticed that I have 9v at BOTH sides of the diode which seemed odd so quickly turned off the system.

 

To be honest I'm not sure which reference point I should be using as the 0v on this test. I tihnk I used GND on the dc power jack, but looking at the Hayato schematic, perhaps I should have used TP19, which is marked as 'zero' (presumably 0v)?? My electronics skills are very rusty!!

 

If there is no current flowing through the diode because Q12 is not on and conducting then there will be no voltage drop so you would read the same voltage both sides of the zenner.

 

I am not sure why TP19 is labelled as 'zero' I cannot see why it would be at zero volts.

Depending of what part of the circuit you are measuring depends on where your 0V reference is, if you place the COM lead or your meter to the 0V drawn the left of MOSFET Q12 then i would expect you will read 9V everywhere, even where the schematic indicates 5V. If you place it on the Com ground drawn to the right of Q12, i.e. the anode of the Zenner or bottom of R75 then I would expect you to read 5V for everything labelled 5V & read 4V for everything labelled as 9V for the reason I mention here.      

 

[bobmoo79]

thanks guys. 

 

 

I checked u6 and it works fine. 

I have now checked all caps and resistors in the power circuit and they're fine. 

I cannot figure out d12 though. I see the reference on the silk screen but the nearby component is not a diode. see below s10 in this 

 

 

 

Next actions: 

1. Remove Q7 from circuit and test with a DVM.
2. Remove Q8 from circuit and test with a DVM.

 

 

Questions for others:

• which component is d12?
• Where is TP15 located? I can't find it anywhere? Is it under the metal shield? (I reinstalled that after the cap replacement so won't see it if it's there).
• Has anybody every noticed an apparent short across R47 before?

Edited August 8, 2019 by bobmoo79

[+Stephen Moss]

14 hours ago, bobmoo79 said:

• which component is d12?
• Where is TP15 located? I can't find it anywhere? Is it under the metal shield? (I reinstalled that after the cap replacement so won't see it if it's there).
• Has anybody every noticed an apparent short across R47 before?

On my Lynx II D12 is immediately to the left of where "D12" is written and immediately to the right of Q12. It might not look like a diode because it is in a SOT23 type package as opposed to the cylindrical MELF package of D9, 10 & 11. If you look closely below it you will see written A & K indicating the Anode and Cathode pins.

 

Looking at the PCB tracks I think TP15 must be under the shield, but you can test that point at the bottom end of R74, immediately above where "Q8" is written.

 

I am not sure why you are testing R47, I cannot locate it on the (Lynx 1) schematics available here but on the Lynx II (Hayato) schematic it is part of Contrast circuit. If you meant R74 which is the dropper for the Zenner then on my Lynx II it measures approximately 117 ohms (in circuit) and R47 measures 55.2K (in circuit) which is what you would expect from resistors that should be 120 (121) and 56K (563) ohms respectively.

 

If you are indeed reading a short across R74 and not R47 then I would suspect Q8 has a base emitter short, I diagnosed a Q8 fault on someone else's board once so it is not unheard of for Q8 to fail. So when you remove Q8 check to see if the short disappears.  If Q8 is shorted possible causes are....

1) Poor circuit design and it is just being over worked - but then you would expect it to be a common fault.

2) A shorted zenner tuning it on to hard

3) Too much current flow due to R73 going low but I would think that unlikely as resistors usually go high resistance/open and you wouldn't be able to draw enough current through U6.    

[Dave222004]

On 8/8/2019 at 10:12 AM, Stephen Moss said:

 

If there is no current flowing through the diode because Q12 is not on and conducting then there will be no voltage drop so you would read the same voltage both sides of the zenner.

 

I am not sure why TP19 is labelled as 'zero' I cannot see why it would be at zero volts.

Depending of what part of the circuit you are measuring depends on where your 0V reference is, if you place the COM lead or your meter to the 0V drawn the left of MOSFET Q12 then i would expect you will read 9V everywhere, even where the schematic indicates 5V. If you place it on the Com ground drawn to the right of Q12, i.e. the anode of the Zenner or bottom of R75 then I would expect you to read 5V for everything labelled 5V & read 4V for everything labelled as 9V for the reason I mention here.      

 

Hi there. Just wandering if you could help. This is in no way contacted to this thread sorry.....

 

I replaced all caps and power regulation components on my lynx 2. I turned it on with the back off and got my finger zapped somewhere around the transormer/inductor area. 

 

The lynx powers up ok but now the screen is very dim all the time.  I'm assuming I've created a short with my finger? The contrast wheel does very very slightly change the contrast when adjusted to where I can barely see the screen. Sound and all buttons work.

 

Could it be a cap that has popped due to everything else seeming ok apart from the screen?

 

Any help would be much appreciated.

 

Thank you.

[+Stephen Moss]

@ Dave 222004, unfortunately, that 2KV transformer for the backlight is right where you fingers goes when holding the Lynx 2 which I also found out the hard way, numbed the end of my finger for about 30 minutes.

 

Not an easy area of the circuit to diagnose problems with as most people would not have access to the required equipment. But what you can do is start by...

1) Look with a magnifying glass around that area for signs of blackening to the PCB or components that appear broken/hole or indentation in them. If there was a short somewhere, particularly of the 2KV then a visual inspection may reveal any resulting damage.

2) You should be able to see the ends of the LCD backlight, is the back light on and if so how bright is it and can you switch it on and off?

3) If you have a multi-meter can you measure the voltage across C23? The schematic indicates it should be around 20V but I cannot provide a comparison measurement as my Lynx is at work I will not be until sometime in September.

 

[Dave222004]

Hi Stephen.

 

Thanks so much for your reply.

 

I have a magnifying glass so will inspect the 2kv area. Hopefully will show something up. The back light does work and can be turned on and off as normal. I will also measure across C23. I will reply back once completed. Thanks again.

 

[battleman13]

The beauty of the lynx 2 is that you can bypass nearly everything that is finicky, and everything that you can bypass isn't necessary for operation.

The ole "5V trick". On a lynx 1 it would be C39, and on a Lynx 2 it would be C41. Simply have a game inserted (or short out pins 31 and 33 on the cart slot) and then supply that capacitor with 5V 1A to the legs, respecting polarity. If your lynx is in working condition, it will work now. If it doesn't work after doing the 5V trick, tracing through the power stage and recapping the unit won't make a difference (save for replacing the two big power caps... that you should do no matter what). So if you can't get the unit working using the 5V trick, I wouldn't worry about anything else. Replace the two largest caps on the board, and then work on getting the game to run with the 5V trick applied. If that trick doesn't work, it means you have damage to either one of the two custom CPU's or damage to the ram chips. As well, it could mean you have a dirty / damaged cart slot. Check to make sure the pins aren't bent or out of place, try a different game if you have one as I have run into the rare game or two that just wouldn't boot  be it they the contacts are dirty and sometimes the cart just outright won't work at all (which is why I prefer to jump pins 31 and 33 on the cart slot, eliminates yet another variable). 

The flex membrane can give you issues, but you've trimmed the very edge of yours... so you should be good there. 

You could have a bad screen, but that is somewhat rare and if it was JUST a bad screen you'd still get sound. Try chips challenge if you have access to it, or crystal mines. They both play sound right away.

The power socket should be replaced, the console 5 kit comes with one. But it's not your issue here.

Console 5 also mentions replacing the part at R74 with a 30 ohm resistor. Supposedly slightly raises voltages a few tenths of a volt, and that helps wake up some picky lynx units.

[+-^CrossBow^-]

On 8/15/2021 at 9:59 PM, battleman13 said:

Console 5 also mentions replacing the part at R74 with a 30 ohm resistor. Supposedly slightly raises voltages a few tenths of a volt, and that helps wake up some picky lynx units.

I've been swapping that resistor out ever since they started to include it. Most Lynx II systems when I've done the McWill upgrades in them, would only read about 4.8v on the +5 after removing the extra stuff required. Replacing out that resistor with the new one will bring it back up to 4.89 or 4.9v in most cases and brings it that much close to the +5 you would expect throughout the system.